Building energy consumption evaluation and scheduling method based on sample entropy

Abstract

The invention discloses a building energy consumption evaluation and scheduling method based on sample entropy, and relates to the field of building energy saving.The method comprises the steps that a plurality of first dimension vectors are formed through building energy consumption time sequences, and the distance between the first dimension vectors is calculated; obtaining the number of the first dimension vectors whose distance is smaller than or equal to a preset threshold; obtaining the number of the second dimension vectors whose distance is smaller than or equal to a preset threshold; calculating a sample entropy based on the number of distances of the building energy consumption time sequence under the first dimension vector and the second dimension vector smaller than a preset threshold value; according to the comparison result of the sample entropy and the sample entropy complexity judgment threshold value, the fluctuation type of the building energy consumption is evaluated, and different scheduling time periods are delimited; and scheduling the building energy consumption by adopting an adaptive scheduling mode. According to the method, the sample entropy serves as a key index, the real fluctuation characteristics of building energy consumption are comprehensively reflected, and a reliable decision basis is provided for energy consumption optimization.

Description

Technical Field

[0001] The present invention relates to the field of building energy conservation, and in particular to a building energy consumption evaluation and scheduling method based on sample entropy. Background Art

[0002] Sample entropy is a measure of the complexity of patterns in a time series. It assesses the randomness and complexity of a time series by calculating the probability of different patterns occurring in the series. Higher sample entropy values ​​indicate a more complex time series, meaning that new, non-repeating patterns are more likely to appear in the series. Lower sample entropy values ​​indicate a more regular and predictable time series, with a higher probability of repeating patterns.

[0003] Introducing the sample entropy time series analysis method into the building energy consumption evaluation process can more accurately evaluate the building energy consumption status and make scientific and reasonable scheduling according to the dynamic changes of energy consumption, so as to improve the building energy utilization efficiency, reduce energy consumption and achieve sustainable development of the construction industry.

[0004] At the same time, traditional building energy consumption evaluation indicators often only focus on the total energy consumption of the building, ignoring the fluctuations in energy consumption in the time dimension and the complex coupling relationship between different energy types. It is difficult to fully and accurately reflect the true characteristics and changing laws of building energy consumption, resulting in the inability to accurately judge the rationality of building energy consumption in practical applications, and unable to provide a scientific and effective basis for energy consumption optimization.

[0005] Furthermore, traditional building energy scheduling strategies rely too heavily on experience and fixed rules, lacking the ability to respond to dynamic changes in building energy consumption in real time and provide intelligent control. When building energy consumption fluctuates abnormally or external environmental conditions change significantly, traditional scheduling methods are unable to adjust the scheduling plan in a timely and flexible manner, resulting in low energy efficiency and making it difficult to achieve energy conservation and emission reduction goals.

[0006] Currently, no effective solutions have been proposed for the problems in related technologies. Summary of the Invention

[0007] In response to the problems in the related art, the present invention proposes a building energy consumption evaluation and scheduling method based on sample entropy to overcome the above-mentioned technical problems existing in the existing related art.

[0008] To this end, the specific technical solutions adopted in the present invention are as follows:

[0009] A building energy consumption evaluation and scheduling method based on sample entropy, comprising:

[0010] Extracting and preprocessing the building energy consumption time series; using the building energy consumption time series to form a number of first-dimensional vectors, and calculating the distance between the first-dimensional vectors; obtaining the number of first-dimensional vectors whose distance is less than or equal to a preset threshold;

[0011] Increasing the dimension of the building energy consumption time series to a second dimension, and obtaining the number of second-dimensional vectors whose distances between vectors are less than or equal to a preset threshold;

[0012] Calculate the sample entropy based on the number of building energy consumption time series whose distances between the first dimension vector and the second dimension vector are less than a preset threshold;

[0013] Based on the comparison results of sample entropy and sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is evaluated and different scheduling periods are delineated; in different scheduling periods, adaptive scheduling methods are used to schedule building energy consumption.

[0014] Furthermore, the extraction and preprocessing of building energy consumption time series include:

[0015] Collect the building energy consumption time series and use the Lagrange difference method to fill the missing values ​​of the building energy consumption time series;

[0016] The Raida criterion is used to remove abnormal values ​​in the building energy consumption time series;

[0017] De-noising of building energy consumption time series based on wavelet transform technology;

[0018] Among them, the building's electricity consumption, building's heat consumption and building's gas consumption are combined to form the building's total energy consumption, and the building's total energy consumption at different times is used to form a building energy consumption time series.

[0019] Furthermore, using the building energy consumption time series to form a number of first-dimensional vectors, and calculating the distance between the first-dimensional vectors includes:

[0020] The building energy consumption time series is composed of several first-dimensional vectors, and the absolute value of the maximum difference between the corresponding elements of the first-dimensional vectors is calculated as the distance between the first-dimensional vectors.

[0021] Furthermore, increasing the dimension of the building energy consumption time series to a second dimension and obtaining the number of second-dimensional vectors whose distances are less than or equal to a preset threshold includes:

[0022] The first dimension is increased to the second dimension, and the distance between the second dimension vectors is calculated, and the number of the second dimension vectors whose distance is less than or equal to a preset threshold is counted, and normalized;

[0023] The method further includes counting the number of second-dimensional vectors whose distances are less than or equal to a preset threshold and normalizing the number of vectors.

[0024] The probability of matching the second dimensional features between the second dimensional vectors under the effective threshold is calculated by using the number of second dimensional vectors whose distances between the normalized second dimensional vectors are less than or equal to the preset threshold.

[0025] Furthermore, based on the number of distances between the building energy consumption time series in the first dimension vector and the second dimension vector being less than a preset threshold, calculating the sample entropy includes:

[0026] Counting the number of first-dimension vectors whose distances are less than or equal to a preset threshold, and normalizing the result;

[0027] Calculate the probability of matching the first dimensional features between the first dimensional vectors under the effective threshold value by using the number of first dimensional vectors whose distances between the normalized first dimensional vectors are less than or equal to the preset threshold value;

[0028] The sample entropy is calculated using the probability of matching the first dimension features between the first dimension vectors under the effective threshold and the probability of matching the second dimension features between the second dimension vectors under the effective threshold.

[0029] Furthermore, the calculation formula of sample entropy is:

[0030]

[0031] Where ampEn(m,r,N) is the sample entropy, m is the dimension, r is the effective threshold, and N is the length of the building energy consumption time series;

[0032] B (m) (r) is the probability that two sequences match the m-dimensional features under the effective threshold r, A (m+1) (r) is the probability that the two sequences match the m+1-dimensional features.

[0033] Furthermore, based on the comparison results of sample entropy and sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is evaluated and different scheduling periods are defined, including:

[0034] Obtaining a sample entropy complexity judgment threshold, and comparing the sample entropy with the sample entropy complexity judgment threshold;

[0035] When the sample entropy is less than the sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is the first type and enters the regular scheduling period;

[0036] When the sample entropy is greater than or equal to the sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is the second type, and it enters the key control and scheduling period.

[0037] Furthermore, obtaining the sample entropy complexity judgment threshold includes:

[0038] Obtain historical data on building energy consumption and calculate sample entropy at different time periods;

[0039] A value in the sample entropy value interval between the first type corresponding time period and the second type corresponding time period is selected as the sample entropy complexity judgment threshold.

[0040] Furthermore, in different scheduling periods, adaptive scheduling methods are used to schedule building energy consumption, including:

[0041] When the scheduling period is the regular scheduling period, based on the feedback control principle, sensors are used to monitor the building environment and equipment status in real time and feed back to the controller;

[0042] Based on the PID control algorithm, the controller adjusts the building equipment according to the control quantity and predicts the deviation trend between the set value and the actual measurement value to optimize the operation of the building equipment.

[0043] Furthermore, in different scheduling periods, the adaptive scheduling method is used to schedule building energy consumption, which also includes:

[0044] When the scheduling period is the key control scheduling period, the target power is calculated based on the energy consumption time series of the building equipment and combined with the performance curve of the building equipment, and converted into variable frequency at the same time;

[0045] The frequency conversion instructions are sent to the inverter to adjust the power of the construction equipment; the monitoring data is fed back through the sensor to form a closed-loop control mechanism.

[0046] The beneficial effects of the present invention are:

[0047] 1. During the data preprocessing phase, this invention uses the Lagrange interpolation method to fill missing values, the 3σ criterion (Laida criterion) to remove outliers, and discrete wavelet transform denoising to improve data quality. Compared with traditional methods, these operations can avoid deviations in building energy consumption analysis caused by missing data, anomalies, or noise interference, laying a solid foundation for subsequent sample entropy-based calculations and evaluations, and improving the credibility and effectiveness of the analysis results.

[0048] 2. The building energy consumption evaluation system, centered on sample entropy, accurately captures the complexity and irregularities of building energy consumption time series. Unlike traditional evaluation indicators that focus solely on total energy consumption, this system considers fluctuations in energy consumption over time and the coupling relationships between different energy types. This fully reflects the true fluctuations in building energy consumption and provides a reliable basis for decision-making regarding energy optimization.

[0049] 3. This invention streamlines the entire process from data processing and energy consumption evaluation to scheduling strategy formulation, enabling intelligent and refined building energy management. Through scientific evaluation and rational scheduling, it can effectively reduce building energy consumption and improve energy efficiency, aligning with the global trend of energy conservation and emission reduction, and providing strong support for the sustainable development of the construction industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 is a flow chart of a building energy consumption evaluation and scheduling method based on sample entropy according to an embodiment of the present invention;

[0052] Figure 2 is a schematic diagram of building energy consumption according to an embodiment of the present invention;

[0053] Figure 3 It is a flow chart of a specific application according to an embodiment of the present invention. DETAILED DESCRIPTION

[0054] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention. They are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. By referring to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0055] According to an embodiment of the present invention, a building energy consumption evaluation and scheduling method based on sample entropy is provided.

[0056] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, the building energy consumption evaluation and scheduling method based on sample entropy according to an embodiment of the present invention includes:

[0057] S1. Extract and preprocess the building energy consumption time series; use the building energy consumption time series to form a number of first-dimensional vectors, and calculate the distance between the first-dimensional vectors; obtain the number of first-dimensional vectors whose distance is less than or equal to a preset threshold.

[0058] S2. Increase the dimension of the building energy consumption time series to a second dimension, and obtain the number of second-dimensional vectors whose distance between the vectors is less than or equal to a preset threshold.

[0059] S3. Calculate sample entropy based on the number of times that the distance between the building energy consumption time series in the first dimension vector and the second dimension vector is less than a preset threshold.

[0060] S4. Based on the comparison results of the sample entropy and the sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is evaluated and different scheduling periods are defined; in different scheduling periods, the building energy consumption is scheduled using an adaptive scheduling method.

[0061] In one embodiment, extracting and preprocessing the building energy consumption time series includes:

[0062] The building energy consumption time series is collected, and the Lagrange difference method is used to fill the missing values ​​of the building energy consumption time series; the Laida criterion is used to eliminate the outliers in the building energy consumption time series; the building energy consumption time series is denoised based on the wavelet transform technology; among them, the building's electricity consumption, building's heat energy consumption and building's gas energy consumption are combined to form the total energy consumption of the building, and the total energy consumption of the building at different times is used to form the building energy consumption time series.

[0063] In one embodiment, using the building energy consumption time series to form a plurality of first-dimensional vectors and calculating the distance between the first-dimensional vectors includes:

[0064] The building energy consumption time series is composed of several first-dimensional vectors, and the absolute value of the maximum difference between the corresponding elements of the first-dimensional vectors is calculated as the distance between the first-dimensional vectors.

[0065] In one embodiment, increasing the dimension of the building energy consumption time series to a second dimension and obtaining the number of second-dimensional vectors whose distances are less than or equal to a preset threshold includes:

[0066] The first dimension is increased to the second dimension, and the distance between the second-dimensional vectors is calculated, and the number of the second-dimensional vectors whose distance is less than or equal to the preset threshold is counted, and normalized.

[0067] Among them, counting the number of second-dimensional vectors whose distances are less than or equal to a preset threshold and normalizing the same also includes: using the number of second-dimensional vectors whose distances are less than or equal to the preset threshold after normalization, calculating the probability of matching the second-dimensional features between the second-dimensional vectors under the effective threshold.

[0068] In one embodiment, based on the number of distances between the first dimension vector and the second dimension vector of the building energy consumption time series being less than a preset threshold, calculating the sample entropy includes:

[0069] The number of first-dimensional vectors whose distances are less than or equal to a preset threshold is counted and normalized; the probability of first-dimensional features matching between first-dimensional vectors under an effective threshold is calculated using the number of first-dimensional vectors whose distances are less than or equal to the preset threshold after normalization; the sample entropy is calculated using the probability of first-dimensional features matching between first-dimensional vectors under an effective threshold and the probability of second-dimensional features matching between second-dimensional vectors under an effective threshold.

[0070] In one embodiment, the sample entropy is calculated as follows:

[0071]

[0072] Where ampEn(m,r,N) is the sample entropy, m is the dimension, r is the effective threshold, and N is the length of the building energy consumption time series; B (m) (r) is the probability that two sequences match the m-dimensional features under the effective threshold r, A (m+1) (r) is the probability that the two sequences match the m+1-dimensional features.

[0073] In one embodiment, based on the comparison result of the sample entropy and the sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is evaluated and different scheduling periods are defined, including:

[0074] Obtain the sample entropy complexity judgment threshold and compare the sample entropy with the sample entropy complexity judgment threshold; when the sample entropy is less than the sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is the first type, and enters the regular scheduling period; when the sample entropy is greater than or equal to the sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is the second type, and enters the key control scheduling period.

[0075] In one embodiment, obtaining the sample entropy complexity determination threshold includes:

[0076] Obtain historical building energy consumption data and calculate sample entropy for different time periods; select a value in the sample entropy value interval between the first type corresponding time period and the second type corresponding time period as the sample entropy complexity judgment threshold.

[0077] In one embodiment, in different scheduling periods, using an adaptive scheduling method to schedule building energy consumption includes:

[0078] When the scheduling period is the regular scheduling period, based on the feedback control principle, sensors are used to monitor the building environment and equipment status in real time and feed back to the controller; based on the PID control algorithm, the controller adjusts the building equipment according to the control quantity and predicts the deviation change trend between the set value and the actual measurement value to optimize the operation of the building equipment.

[0079] In one embodiment, in different scheduling periods, scheduling building energy consumption using an adaptive scheduling method further includes:

[0080] When the scheduling period is the key control scheduling period, the target power is calculated based on the energy consumption time series of the building equipment and combined with the performance curve of the building equipment, and converted into variable frequency at the same time; the variable frequency instruction is sent to the frequency converter to adjust the power of the building equipment; the monitoring data is fed back through the sensor to form a closed-loop control mechanism.

[0081] In order to facilitate understanding of the above technical solutions of the present invention, the working principle of the present invention in actual process is described in detail below.

[0082] like Figure 2 As shown in the building energy consumption diagram, building energy consumption can be categorized based on three loads: electricity, heat, and gas. The electrical load includes the lighting system (electricity used by various indoor and outdoor lighting fixtures), the air conditioning system (electricity used by central air conditioners, split air conditioners, and other cooling and heating systems), the ventilation system (electricity used by fans and exhaust fans), the elevator system (electricity used by elevators and escalators), the fire protection system (electricity used by fire pumps, emergency lighting, etc.), electrical appliance consumption (electricity used by office and home computers, printers, televisions, etc.), and the water supply and drainage system (electricity used by equipment such as water pumps). The thermal load includes heating energy consumption (energy consumed by heating systems such as radiators and floor heating in winter), hot water supply energy consumption (heating energy consumed to meet domestic hot water needs), air conditioning energy consumption (heat consumption consumed by air conditioners for heating or cooling), and cooking energy consumption (heat consumption generated by kitchen cooking equipment). Gas load involves heating energy consumption (gas consumption of gas-fired heating equipment such as gas boilers), hot water supply energy consumption (gas consumption of gas-fired water heaters and other equipment supplying hot water), absorption air conditioning energy consumption (gas consumption of gas-fired absorption air conditioners) and cooking energy consumption (gas consumption of gas-fired stoves, ovens and other cooking equipment).

[0083] like Figure 3 , the present invention comprises:

[0084] Step 1: Obtain the collected time series of building energy consumption and perform preprocessing.

[0085] Step 2: Convert the time series Y m (i) Form an m-dimensional vector and calculate the vector Y m (i) and other vectors Y m (j) The distance between them.

[0086] Step 3: Count the given Y m (i) and Y m (j) The number of intervals less than the threshold, then increase the dimension to m+1, and count Y m+1 (i) and Y m+1 (j) corresponding number.

[0087] Step 4: Calculate the sample entropy and evaluate the scheduling methods for different scheduling periods based on the sample entropy. That is, calculate the sample entropy based on the processed time series, and determine the scheduling period and the scheduling methods for different scheduling periods based on the sample entropy.

[0088] The details are as follows:

[0089] Step 1: Obtain the collected time series of building energy consumption and perform preprocessing, including:

[0090] (1) Fill in missing values ​​in time series:

[0091]

[0092] In this invention, the Lagrange difference method is used to fill the missing values ​​of the building energy consumption time series. Where t is the missing time point in the time series, X(t) is the missing value in the time series, and t i is the time of the known data point, X(t i ) is the corresponding time series value, t j is the time of another known data point, X(t j ) is another corresponding time series value, and n is the time series. This formula is used to effectively fill in missing values ​​and ensure data integrity.

[0093] (2) Fill in and remove outliers in the time series:

[0094]

[0095] In the present invention, the 3σ criterion is adopted to remove abnormal values ​​in the time series. is the mean of the time series, and σ is the variance of the time series. The range of abnormal values ​​is arrive If a data point X(t) satisfies or Then X(t) can be determined to be an outlier and removed to improve the accuracy of the data.

[0096] (3) Time series denoising:

[0097]

[0098] First, the time series is subjected to discrete wavelet transform to decompose it into wavelet coefficients of different scales and positions. j，k is the wavelet coefficient, j is the scale, k is the position, ψ j，k (t i ) is a wavelet basis function obtained by scaling and translating the basic wavelet function.

[0099]

[0100] Then the wavelet threshold is processed. is the processed wavelet coefficient, λ is the threshold, S is the noise standard deviation, Sgn is the sign function, when the parameter is positive, the Sgn function returns 1; when it is negative, it returns -1; and when the parameter is 0, it returns 0.

[0101]

[0102] where ψ' j，k (t i ) is the reconstructed wavelet function. Finally, wavelet reconstruction is performed to obtain the processed time series Enhance data availability.

[0103] Step 2: Convert the time series Y m (i) Form an m-dimensional vector and calculate the vector Y m (i) and other vectors Y m (j) The distance between includes:

[0104] (1) Extract the building energy consumption time series and convert the time series Y with a fixed length of N into m (i) Form an m-dimensional vector:

[0105] X(t)=X e (t)+X h (t)+X g (t);

[0106] Where X(t) is the total energy consumption of the building at time t, X e (t) is the electricity consumption of the building at time t, X h (t) is the heat energy consumption of the building at time t, X g (t) is the gas energy consumption of the building at time t.

[0107] Y m (i)={X(i),

[0108] X(i) is the total energy consumption of the building at the i-th moment, X(i+1) is the total energy consumption of the building at the i+1-th moment, and X(i+m-1) is the total energy consumption of the building at the i+m-1-th moment. These vectors represent the m consecutive values ​​of X starting from the i-th point.

[0109] (2) Calculate vector Y m (i) and other vectors Y m (j) The distance between:

[0110] d[Y m (i), Ym (j)]=max k＝0，1，...，m-1 (|X(i+k)-X(j+k)|);

[0111] where d[Y m (i), Y m (j)] is the vector Y m (i) and vector Y m The distance between (j) is the absolute value of the maximum difference between the corresponding elements of the two.

[0112] Step 3: Count the given Y m (i) and Y m (j) The number of intervals less than the threshold, then increase the dimension to m+1, and count Y m+1 (i) and Y m+1 (j) The corresponding number includes:

[0113] (1) Calculate for a given Y m (i), statistics Y m (i) and Y m (j) The number of distances between them is less than or equal to the set threshold r;

[0114]

[0115] For a given Y m (i), statistics Y m (i) and Y m The number of j (1<<j<<N-m+1, j≠i) whose distance between them is less than or equal to r, and is recorded as B i .

[0116] definition:

[0117]

[0118] in B i The normalized result, B (m) (r) is the probability that two sequences match the m-dimensional features under the effective threshold r.

[0119] (2) Increase the dimension to m+1 and calculate Y m+1 (i) and Y m+1 (j) the number of distances less than or equal to the effective threshold r;

[0120]

[0121] For a given Y m+1 (i), statistics Y m+1 (i) and Y m+1The number of j (1<<j<<Nm, j≠i) whose distance between them is less than or equal to r, and is recorded as A i .

[0122] definition:

[0123]

[0124] in A i The normalized result of A (m+1) (r) is the probability that two sequences match the m+1-dimensional features under the effective threshold r.

[0125] Step 4: Calculate the sample entropy and evaluate the scheduling methods for different scheduling periods based on the sample entropy.

[0126] When the time series Y m When the length N of (i) is finite, the sample entropy can be defined as:

[0127]

[0128] When the time series Y m When the length N of (i) is infinite, the sample entropy can be defined as:

[0129]

[0130] Among them B (m) (r) is the probability that two sequences match the m-dimensional feature under the effective threshold r, and A (m+1) (r) is the match between two sequences m +Probability of 1-dimensional features.

[0131] Because the sample entropy is directly related to the embedding dimension m and the effective threshold r, these two values ​​must be determined first. According to research results, m = 1 & 2, r = 0.1-0.25σ (σ is the standard deviation of the original data), the calculated sample entropy has relatively reasonable statistical characteristics. In this invention, to prevent the loss of too much important information, m is set to 2. At the same time, r that is too large or too small will cause the sample entropy to fail to truly reflect the situation. Therefore, through analysis and comparison, r = 0.20σ in this invention.

[0132] The fluctuation types of building energy consumption are evaluated and classified based on the relationship between the sample entropy SampEn(m, r, N) and the sample entropy complexity judgment threshold TH. When SampEn(m, r, N) < TH, it indicates that the energy consumption data fluctuates relatively smoothly and can be designated as a regular scheduling period. When SampEn(m, r, N) ≥ TH, it means that the energy consumption fluctuates violently and is designated as a key control scheduling period. Differentiated scheduling methods are adapted for different scheduling periods. During the regular scheduling period, a rule-based scheduling strategy is adopted to maintain stable equipment operation. During the key control scheduling period, the peak-shaving capacity of the building energy supply system is enhanced, and the equipment operating parameters are fine-tuned to achieve the goal of reducing energy consumption and improving energy utilization efficiency.

[0133] The threshold TH is primarily determined by analyzing historical building energy consumption data and calculating sample entropy values ​​for different time periods. The difference in sample entropy values ​​between periods of significant energy consumption fluctuations and relatively stable periods is observed, and a value between the sample entropy values ​​for stable periods and periods of significant fluctuations is selected as the threshold TH. In this invention, the threshold TH ranges from 0.6 to 0.7.

[0134] During regular scheduling periods, a rule-based scheduling strategy is adopted to maintain stable equipment operation. The main methods include: using an intelligent control system based on the feedback control principle, monitoring the building environment and equipment status in real time through sensors, and feeding back information to the controller. The controller uses the set PID control algorithm and the deviation between the set value and the actual measured value to obtain the control quantity to adjust the equipment through the three links of proportional (P), integral (I), and differential (D). The proportional link quickly responds to deviations, the integral link eliminates static deviations, and the differential link predicts the trend of deviation changes. Automatically optimize equipment operation, achieve precise control, and improve equipment operation stability and energy efficiency. Strengthening equipment maintenance, regular inspection and replacement of components can avoid overall performance degradation or failure due to local failures. Reduce equipment operation resistance, enable equipment to operate in an efficient state, and reduce energy consumption and operational instability factors.

[0135] During key control and scheduling periods, the peak-shaving capacity of the building energy supply system is enhanced, and equipment operating parameters are finely regulated to reduce energy consumption and improve energy efficiency. Key methods include optimizing equipment operating modes. For large energy-consuming equipment such as cooling and heating, target power is calculated based on the equipment's energy consumption time series and combined with the equipment's performance curve. This target power is then converted into a variable frequency drive (VFD). This frequency command is then sent to the VFD, which adjusts the output frequency to drive the motor, thereby changing the equipment's power. During operation, sensor feedback is used. If deviations exceed a certain range, the frequency is readjusted, forming a closed-loop control loop. This achieves variable frequency speed regulation to match output power to actual load, reducing energy waste. Energy storage devices, such as lithium-ion batteries and flow batteries, are integrated into the energy supply system and equipped with appropriate charge and discharge control devices. During periods of low energy consumption, the energy storage devices charge and store energy. During peak consumption, they discharge and release energy to supplement the energy supply system's shortfall, thus shaving peaks and filling valleys and smoothing energy consumption fluctuations.

[0136] In summary, with the help of the above technical solutions of the present invention, the present invention uses sample entropy as a key indicator to construct a scientific and comprehensive building energy consumption evaluation system. This system can accurately capture the complexity and irregularity of building energy consumption time series, comprehensively reflect the true fluctuation characteristics of building energy consumption, and provide a reliable decision-making basis for energy consumption optimization. By calculating sample entropy and dividing the scheduling period according to its relationship with the threshold, differentiated scheduling methods are adapted for different scheduling periods. In the regular scheduling period, a rule-based scheduling strategy is adopted. In the key control scheduling period, the peak-shaving capacity of the building energy supply system is enhanced and the equipment operating parameters are fine-tuned. This realizes the intelligent and refined management of building energy consumption, effectively reduces building energy consumption, and improves energy utilization efficiency.

[0137] During the data preprocessing phase, this invention utilizes the Lagrangian interpolation method to fill missing values, the 3σ criterion to eliminate outliers, and the discrete wavelet transform for denoising, effectively improving data quality through a multi-pronged approach. Compared to traditional methods, these operations can avoid bias in building energy consumption analysis caused by missing data, anomalies, or noise interference, laying a solid foundation for subsequent calculation and evaluation based on sample entropy, and enhancing the credibility and validity of the analysis results. This building energy consumption evaluation system, centered on sample entropy, accurately captures the complexity and irregularities of building energy consumption time series. Unlike traditional evaluation indicators that focus solely on total energy consumption, this invention considers the fluctuations in energy consumption over time and the coupling relationships between different energy types, comprehensively reflecting the true fluctuation characteristics of building energy consumption and providing a reliable basis for decision-making on energy consumption optimization. This invention streamlines the entire process from data processing, energy consumption evaluation, to scheduling strategy formulation, achieving intelligent and refined building energy management. Through scientific evaluation and rational scheduling, building energy consumption can be effectively reduced and energy efficiency improved, aligning with the global trend of energy conservation and emission reduction and providing strong support for the sustainable development of the construction industry.

[0138] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A building energy consumption evaluation and scheduling method based on sample entropy, characterized in that: include: Extract and preprocess building energy consumption time series; Using the building energy consumption time series to form a number of first-dimensional vectors, and calculating the distance between the first-dimensional vectors; Obtaining the number of first-dimensional vectors whose distance is less than or equal to a preset threshold; Increasing the dimension of the building energy consumption time series to a second dimension, and obtaining the number of second-dimensional vectors whose distances between vectors are less than or equal to a preset threshold; Calculate the sample entropy based on the number of building energy consumption time series whose distances between the first dimension vector and the second dimension vector are less than a preset threshold; Based on the comparison results of sample entropy and sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is evaluated and different scheduling periods are delineated; in different scheduling periods, adaptive scheduling methods are used to schedule building energy consumption.

2. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 1 is characterized in that: The extraction and preprocessing of building energy consumption time series includes: Collect the building energy consumption time series and use the Lagrange difference method to fill the missing values ​​of the building energy consumption time series; The Raida criterion is used to remove abnormal values ​​in the building energy consumption time series; De-noising of building energy consumption time series based on wavelet transform technology; Among them, the building's electricity consumption, building's heat consumption and building's gas consumption are combined to form the building's total energy consumption, and the building's total energy consumption at different times is used to form a building energy consumption time series.

3. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 1 is characterized in that: The method of forming a plurality of first-dimensional vectors by using the building energy consumption time series and calculating the distance between the first-dimensional vectors includes: The building energy consumption time series is composed of several first-dimensional vectors, and the absolute value of the maximum difference between the corresponding elements of the first-dimensional vectors is calculated as the distance between the first-dimensional vectors.

4. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 1 is characterized in that: Increasing the dimension of the building energy consumption time series to a second dimension and obtaining the number of second-dimensional vectors whose distances are less than or equal to a preset threshold includes: The first dimension is increased to the second dimension, and the distance between the second dimension vectors is calculated, and the number of the second dimension vectors whose distance is less than or equal to a preset threshold is counted, and normalized; The method further includes counting the number of second-dimensional vectors whose distances are less than or equal to a preset threshold and normalizing the number of vectors. The probability of matching the second dimensional features between the second dimensional vectors under the effective threshold is calculated by using the number of second dimensional vectors whose distances between the normalized second dimensional vectors are less than or equal to the preset threshold.

5. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 1 is characterized in that: The calculating of the sample entropy based on the number of distances between the first dimension vector and the second dimension vector of the building energy consumption time series that are less than a preset threshold comprises: Counting the number of first-dimension vectors whose distances are less than or equal to a preset threshold, and normalizing the result; Calculate the probability of matching the first dimensional features between the first dimensional vectors under the effective threshold value by using the number of first dimensional vectors whose distances between the normalized first dimensional vectors are less than or equal to the preset threshold value; The sample entropy is calculated using the probability of matching the first dimension features between the first dimension vectors under the effective threshold and the probability of matching the second dimension features between the second dimension vectors under the effective threshold.

6. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 5 is characterized in that: The calculation formula of the sample entropy is: Where ampEn(m,r,N) is the sample entropy, m is the dimension, r is the effective threshold, and N is the length of the building energy consumption time series; B (m) (r) is the probability that two sequences match the m-dimensional features under the effective threshold r, A (m+1) (r) is the probability that the two sequences match the m+1-dimensional features.

7. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 1 is characterized in that: The comparison result between the sample entropy and the sample entropy complexity judgment threshold is used to evaluate the fluctuation type of building energy consumption and define different scheduling periods, including: Obtaining a sample entropy complexity judgment threshold, and comparing the sample entropy with the sample entropy complexity judgment threshold; When the sample entropy is less than the sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is the first type and enters the regular scheduling period; When the sample entropy is greater than or equal to the sample entropy complexity judgment threshold, the fluctuation type of building energy consumption is the second type, and it enters the key control and scheduling period.

8. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 7 is characterized in that: The method of obtaining the sample entropy complexity judgment threshold comprises: Obtain historical data on building energy consumption and calculate sample entropy at different time periods; A value in the sample entropy value interval between the first type corresponding time period and the second type corresponding time period is selected as the sample entropy complexity judgment threshold.

9. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 7 is characterized in that: The method of scheduling building energy consumption in different scheduling periods using an adaptive scheduling method includes: When the scheduling period is the regular scheduling period, based on the feedback control principle, sensors are used to monitor the building environment and equipment status in real time and feed back to the controller; Based on the PID control algorithm, the controller adjusts the building equipment according to the control quantity and predicts the deviation trend between the set value and the actual measurement value to optimize the operation of the building equipment.

10. The building energy consumption evaluation and scheduling method based on sample entropy according to claim 7 is characterized in that: The method of scheduling building energy consumption in different scheduling periods using an adaptive scheduling method also includes: When the scheduling period is the key control scheduling period, the target power is calculated based on the energy consumption time series of the building equipment and combined with the performance curve of the building equipment, and converted into variable frequency at the same time; The frequency conversion instructions are sent to the inverter to adjust the power of the construction equipment; the monitoring data is fed back through the sensor to form a closed-loop control mechanism.

Images